Cloning and characterization of complementary DNA encoding human N-acetylglucosamine-phosphate mutase protein

Gene 242 (2000) 97–103 www.elsevier.com/locate/gene

Cloning and characterization of complementary DNA encoding human N-acetylglucosamine-phosphate mutase protein Chaoyang Li, Marilis Rodriguez, Debendranath Banerjee * Department of Membrane Biochemistry II, The Lindsley F. Kimball Research Institute, New York Blood Center, New York, NY 10021, USA Received 23 September 1999; accepted 29 November 1999 Received by D. Schlessinger

Abstract Endothelial cells express erythropoietin receptor (EpoR) and are responsive to erythropoietin (Epo). Upon ligand binding, EpoR activates multiple signaling cascades. Identification of genes expressed in response to Epo is important for understanding the molecular nature of the signals. Applying the differential display approach, an effective method for analysis of gene expression, we identified five differentially expressed mRNAs. In this study, we cloned human N-acetylglucosamine-phosphate mutase from a human microvascular endothelial cell (HMVEC ) cDNA library using one of the differentially expressed fragments as a probe. The nucleotide (nt) sequence analysis of the longest clone displayed a 2 kb cDNA fragment and encodes a protein of ~542 amino acids with a predicted MW of ~60 kDa. Northern blotting and reverse transcriptase-polymerase chain reaction analysis revealed an upregulation of the N-acetylglucosamine-phosphate mutase mRNA after 2 h of stimulation of cells with Epo. This gene was shown to be variably expressed in human tissues and is located on chromosome 6. These studies demonstrate that the expression of N-acetylglucosamine-phosphate mutase mRNA responds to cytokines, and the presence of a 10 aa motif similar to the putative active site of several hexose-phosphate mutases provides a basis for future studies of the role of this gene in the regulation of Epo-stimulated endothelial cell proliferation. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Differential display; Endothelial cells; Erythropoietin; Hexose-phosphate mutase; Northern blot; Nucleotide sequence; Recombinant DNA

1. Introduction Erythropoietin ( Epo) is a glycoprotein hormone that functions as a major regulator of erythropoiesis. In mammals, Epo is produced in fetal liver and adult kidney ( Koury and Bondurant, 1992). It is well established that Epo gene expression is controlled by oxygen tension. Hypoxic and anemic conditions result in elevated Epo expression (Jelkmann, 1992). Until recently,

Abbreviations: aa, amino acid(s); AGM1, human N-acetylglucosamine-phosphate mutase gene; bp, base pair(s); cDNA, complementary DNA; dd, differential display; dNTP, deoxyribonucleoside triphosphate; Epo, erythropoietin; EpoR, erythropoietin receptor; HMVEC, human microvascular endothelial cell(s); HUVEC, human umbilical vein endothelial cell(s); kb, kilobase(s) or 1000 bp; NH , amino termi2 nal; nt, nucleotide(s); ORF, open reading frame; PCR, polymerase chain reaction; RT, reverse transcriptase; UDP-GlcNAc, 5∞-uridinediphospho-N-acetyl--glucosamine; UNT, untranslated nucleotides. * Corresponding author. Tel.: +1-212-570-3205; fax: +1-212-879-0243. E-mail address: [email protected] (D. Banerjee)

erythroid precursor cells have been thought to be the exclusive target for Epo. However, recent reports suggest that Epo may play important roles in endothelial cell functions. Epo functions through its interaction with a single chain cell surface receptor of the cytokine receptor superfamily (Bazan, 1990; D’Andrea et al., 1989). In hematopoietic cells, EpoR mRNA is expressed at moderate levels (Orlic et al., 1995). EpoR or EpoR mRNA are also expressed in non-hematopoietic cells, including human umbilical vein endothelial cells ( HUVEC ) (Anagnostou et al., 1994), rat brain capillary endothelial cells ( Yamaji et al., 1996), murine hippocampal and cerebrocortical areas (Morishita et al., 1997) and primary cultured hippocampal and cortical neurons (Morishita et al., 1996; Digicaylioglu et al., 1995). The EpoRs are functional in HUVEC and mouse brain. Epo stimulates proliferation and migration of human and bovine endothelial cells and also angiogenesis of rat thoracic aorta (Anagnostou et al., 1990; Carlini et al., 1995). The angiogenic effect of Epo has been studied in

0378-1119/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S0 3 7 8 -1 1 1 9 ( 9 9 ) 0 0 54 3 - 0

98

C. Li et al. / Gene 242 (2000) 97–103

ovariectomized mice, where injection of Epo into the uterine cavity promoted blood vessel formation in the endometrium ( Yasuda et al., 1998). Angiogenesis is a complex process requiring migration of endothelial cells, their proliferation, and their differentiation into tube-like structures (Risau, 1997). Although considerable attention has been focused on the mechanisms involved in the regulation of endothelial cell growth, little is known about the molecular events associated with the organisation/differentiation of endothelial cells into capillaries (Maciag et al., 1982). The endothelial cell is capable of activating a unique genetic program in response to environmental signals (Risau, 1995). Among others, interleukin-1, c-interferon and bacterial endotoxin inhibit endothelial growth and promote a morphological change that resembles the early stage of differentiation (Maciag et al., 1982). Furthermore, culture on three-dimensional gels also induces endothelial differentiation and capillary formation (Ingber and Folkman, 1989). Since Epo affects endothelial cell function, we reasoned that the isolation of differentially expressed genes in Epo-treated endothelial cells would yield insights into the molecular mechanisms achieving a better understanding of normal endothelial cell function and differentiation. Here, we report the cloning, sequencing, and characterization of the first of a set of genes shown by differential display to be upregulated by Epo treatment in human microvascular cells. It is AGM1, the human homologue of the yeast N-acetylglucosamine-phosphate mutase gene.

2. Materials and methods 2.1. Cell culture and erythropoietin treatment Primary cultures of human microvascular endothelial cells (HMVEC ) were purchased from Clonetics (San Diego, CA) and used between passages 2 and 4. Cells were cultured at 37°C in modified MCDB 131 medium ( EBM ) supplemented with 5% fetal bovine serum (Hyclone Laboratories, Logan, UT ) in an air–5% CO 2 atmosphere at constant humidity. Human recombinant erythropoietin (Amgen, CA) was diluted to a concentration of 10 mg/ml with Iscove’s modified Dulbecco’s medium ( IMDM ) containing 1% bovine serum albumin (BSA). The diluted erythropoietin was added to the culture medium (to a final concentration of 52 ng/ml ) of a confluent monolayer of endothelial cells. After 24 h of incubation, the medium was removed, and cells were utilized for RNA preparation. 2.2. RNA isolation and differential display Total cellular RNA from 80% confluent cultures of endothelial cell monolayers grown either in the absence

or presence of Epo was extracted by TRIzol (Life Technologies, NY ), according to the manufacturer’s instructions. After removal of chromosomal DNA contamination in total RNAs using the MessageClean kit (GenHunter, TN ), a differential display was carried out by using H-T11M (M=G, A, or C ) and H-AP1 (5∞-AAGCTTGATTGCC-3∞) arbitrary primers and the RNA-map kit (GenHunter) as instructed by the manufacturer. The differentially displayed cDNA fragment from a dried sequencing gel was recovered essentially as described by Liang and Pardee (1992). 2.3. Reamplification, cloning and sequencing of the differentially expressed band The cDNA fragment was PCR reamplified. The primer set and the PCR conditions were the same as described above except the dNTP concentrations were increased to 20 mM, and the radioactive nt was omitted from the reaction mixture. The amplified product was checked on a 2.0% agarose gel, and the DNA fragment was cloned into the plasmid pCR 2.1 using a TA cloning Kit (Invitrogen, CA). Several well-isolated white colonies were picked and grown overnight. The plasmid DNA was prepared using QIAprep Spin Miniprep Kit (QIAGEN, CA), and appropriate recombinants were selected for double-stranded DNA sequencing. 2.4. cDNA cloning and sequencing A cDNA library was prepared from erythropoietintreated HMVEC poly(A)+ RNA and cloned into a l-UniZAP XR vector (Stratagene, La Jolla, CA). The library was screened with a 32P-labeled dd DNA fragment according to the method described previously (Bhattacharyya et al., 1991). From several positive plaques, the pBluescript phagemids were excised by the in-vivo excision protocol using ExAssist helper phage and E. coli SOLR strain (Stratagene). The phagemid DNAs were purified using QIAprep Spin Miniprep Kit. The insert DNAs were sequenced in both directions on an ABI 373 DNA sequencer with ABI Prism kits (Applied Biosystems Division of Perkin-Elmer, Foster City, CA). 2.5. RNA and DNA blot analysis Poly(A)+ RNA was isolated from total RNA using Oligotex mRNA kit (QIAGEN ), according to the supplier’s instructions. Poly(A)+ RNAs were run on a formaldehyde /agarose gels containing 0.66 M formaldehyde, 40 mM MOPS-NaOH (pH 7.2), 10 mM sodium acetate, and 1 mM EDTA and blotted onto a Hybond-N+ membrane (Amersham, IL). The blots were hybridized with a 32P-labeled cDNA insert in QuickHyb (Stratagene, CA) at 68°C for 2 h. The blots

C. Li et al. / Gene 242 (2000) 97–103

were washed with 0.1× SSC containing 0.1% SDS at 60°C and exposed on X-ray film. A monochromosomal somatic cell hybrid blot, containing 5 mg of DNA samples, digested with EcoRI, from 24 somatic hybrid cell lines plus three control DNAs was obtained from Quantum Technologies (Montreal,Canada). The blot was hybridized overnight with a 32P-labeled AGM1 cDNA fragment as a probe in ExpressHyb (Clontech, Palo Alto, CA) at 65°C and washed with 0.1× SSC containing 0.5% SDS at 55°C and autoradiographed.

3. Results 3.1. Cloning and cDNA sequencing of human AGM1 To identify genes expressed as a consequence of Epo treatment, mRNA populations from normal and Epotreated human microvascular endothelial cells, cultured under the same conditions, were compared by the differential display technique. Complementary DNA fragments corresponding to five apparently differentially expressed mRNAs were recovered and cloned. The characterization of one of these clones is the subject of this report. The differentially expressed band was recovered from the dried gel and reamplified using the corresponding T11G and 5∞-AAGCTTGATTGCC-3∞ primer set. The cDNA product (~210 bp) was consistent in size with the original displayed gel ( Fig. 1A). The cDNA fragment was cloned into pCR2.1 vector and sequenced. Fig. 1B shows the nucleotide sequence. The 5∞ end of the sequence matched perfectly with the primer used, while the 3∞ end showed ~75% match with T11G primer. The sequence appears to be AT-rich, with a putative polyadenylation signal located close upstream from the 3∞end. As expected, the nucleotide sequence analysis suggests that the cDNA fragment corresponds to the 3∞ end of the mRNA. To obtain a full-length clone, the initial 210 bp fragment was used as a probe to screen an endothelial cell cDNA library. From approximately 1×106 plaques,

99

eight positives were obtained. After two rounds of purification of the phages, five individual clones were isolated by in-vivo excision, and their sequences were determined. The sequence of the longest clone is shown in Fig. 2A. The complete nucleotide sequence of the cDNA is 1989 bp long and contains a continuous ORF of 1629 bp with an in-frame stop codon 21 bp upstream of the putative start codon. The first ATG codon lies within a consensus translation initiation sequence, i.e. CGAGACATGG, which partially fulfills the optimal sequence for a eukaryotic translation start site ( Kozak, 1991). The 5∞ UNT has a high GC content of 65%, with two CpG dinucleotides within 40 bp. The relatively large 3∞ UNT ( 320 bp) contained two ATTTA motifs associated with mRNA instability (Shaw and Kamen, 1986). A polyadenylation signal, ATTAAA, was detected 11 nt upstream of the poly(A) tail. The ORF encoded a protein of ~59 620 Da consisting of 542 amino acids. The deduced amino acid sequence was highly hydrophilic, with no hydrophobic NH -terminal region ( Fig. 2B). A sequence comparison 2 by the FASTA and BLAST programs revealed identity with human N-acetylglucosamine-phosphate mutase protein (GenBank Accession No. AF 102265). Starting at aa residue 58, a short region of 10 amino acids (GVMVTASHNP) showed a high similarity to the putative active site motif of several hexose phosphate mutases. Thus, the cDNA clone was named AGM1. 3.2. AGM1 mRNA expression in human endothelial cells To confirm stimulation of AGM1 mRNA in response to erythropoietin, a Northern blot analysis was performed using [32P]dATP-labeled AGM1 cDNA as a probe. As demonstrated in Fig. 3A, stimulation of the HMVECs with Epo resulted in an increase in AGM1 mRNA. The findings obtained by the Northern blotting approach were confirmed by RT-PCR using specific AGM1 primers for detection of 700 bp AGM1 cDNA. As shown in Fig. 3B, stimulation of HMVECs with Epo resulted in a marked increase of the ~700 bp AGM1 cDNA.

Fig. 1. Reamplification, cloning and sequencing of the differential display DNA fragment. The differentially expressed cDNA fragment was PCRreamplified using the same primer set and the PCR conditions and cloned as described in Section 2. (A) Agarose gel electrophoresis. An aliquot of amplified product was separated on a 2.0% agarose gel ( lane 2) along with 100 bp molecular size marker ( lane 1). (B) Nucleotide sequences. The plasmid DNA was prepared from appropriate recombinants, and double-stranded DNA sequencing was performed. The nt of the coding strand are shown. The flanking mRNA mapping primers are underlined.

100

C. Li et al. / Gene 242 (2000) 97–103

A

B

C. Li et al. / Gene 242 (2000) 97–103

Fig. 3. Effect of Epo on AGM1 mRNA expression in primary culture of HMVEC. (A) Northern blot. Confluent monolayers of HMVEC were incubated in the absence or presence of Epo (52 ng/ml ). After Epo exposure for the prescribed times, the cells were washed, and poly(A)+ RNA was isolated as described in Section 2. RNAs were resolved on a 1% denaturing agarose gel, blotted, hybridized and exposed on X-ray film for 72 h at −80°C. RNA size markers, human 28 S (4.8 kb) and 18 S (1.9 kb) ribosomal RNA, were used to calculate the size of AGM1 mRNA. (B) RT-PCR. The procedure for PCR amplification was as described in the legend of Fig. 1A. The following primer sets were used for AGM1, 5∞-CATGGATTTAGGTG-3∞and 5∞-TTCCCTTAGCTCAGGGCC-3∞; for glyceraldehyde-3-phosphate dehydrogenase (GAPDH ), 5∞-ACCACAGTCCATGCCATCAC-3∞ and 5∞-TCCACCACCCTGTTGCTGTA-3∞ (Clontech). The upper panel shows ethidium staining of the AGM1 amplified product (~700 bp), and the lower panel shows the GAPDH product (~452) used as a RNA input control.

3.3. AGM1 expression in human tissues The expression of AGM1 mRNA in various human adult tissues was determined by Northern blot analysis using an AGM1 cDNA probe. With the exception of lung, AGM1 mRNA expression (~2.4 kb) was found at different levels in the majority of human adult tissues examined ( Fig. 4A). When normalized to actin, the AGM1 mRNA expression levels were relatively high in pancreas, heart, liver, and placenta, and relatively low in brain, skeletal muscle and kidney. In all tissues, the major mRNA species was ~2.4 kb, and two larger

101

Fig. 4. AGM1 mRNA expression in human tissues. (A) The mRNA blot containing human adult tissue mRNAs was obtained from Clontech Laboratories and was hybridized with a 32P-labeled AGM1 cDNA insert. The prehybridization and hybridization were in ExpressHyb (Clontech) at 68°C for 2 h. The blots were washed with 0.1× SSC containing 0.1% SDS at 55°C and exposed on X-ray film. (B) The same blot was stripped and rehybridized with a 32P-labeled b-actin probe. Hybridization to b-actin indicates that similar amounts of RNA were used per lane; as indicated by the manufacturer, heart and skeletal muscle contain two forms of b-actin, 2.0 and 1.8 kb.

transcripts of ~5 and 8 kb were noted. In the pancreas, three additional bands were also observed. The hybridization profile of actin is shown in Fig. 4B. 3.4. Mapping of the human AGM1 gene to chromosome 6 A somatic cell hybrid panel consisting of hamster, human, and mouse DNA was digested with EcoRI, separated on a agarose gel and blotted (Quantum Technologies, Montreal, Canada). Probing the Southern blot with a human AGM1 cDNA probe revealed one specific band in the human genomic DNA, and the somatic hybrid cell line 6, which contains human chromosome 6 ( Fig. 5; lane 6). All other hybrid cell lines were negative for the human-specific band. Thus, the human AGM1 gene maps to human chromosome 6.

4. Discussion Angiogenesis is a process of central importance during embryonic development, for tissue regeneration,

Fig. 2. (A) cDNA sequence of the human AGM1 gene and deduced aa sequences (GenBank Accession No. AF 180371). The recombinant phagemid DNAs were purified. The insert DNAs were sequenced in both directions, as described in Section 2. The nt sequence of the coding strand is shown from the 5∞3∞ direction. The nt and aa positions are numbered. The presumed translation start codon at nt 1–3, in-frame stop codon, potential mRNA destabilizing elements (ATTTA) and the polyadenylation signal (ATTAAA) are underlined. The termination codon at nt 1627–1629 is marked with triple asterisks. The dd nts are boxed, and the putative 10 aa hexosephosphate mutase motif is underlined. (B) Hydropathic profile of the deduced aa of AGM1. The hydrophobicity values were obtained according to the algorithm of Kyte and Doolittle (1982) with a window of 5 aa residues. Positive values represent increased hydrophobicity.

102

C. Li et al. / Gene 242 (2000) 97–103

Fig. 5. Mapping of AGM1 to human chromosome 6. A monochromosomal somatic cell hybrid blot, containing 5 mg of DNA samples, digested with EcoRI, from 24 somatic hybrid cell lines plus three control DNAs was obtained from Quantum Technologies (Montreal,Canada). The blot was probed with a 32P-labeled human AGM1 cDNA fragment. An autoradiogram is shown. The human-specific bands are shown in control lanes and in lane 6, which corresponds to human chromosome 6.

and in the progression of tumors ( Risau, 1997). We have cloned and characterized a N-acetylglucosaminephosphate mutase gene whose expression pattern in response to Epo indicates that it is involved in the signal transduction pathway that ultimately leads to proliferation of endothelial cells and formation of new blood vessels. N-acetylglucosamine-phosphate mutase is involved in the reversible interconversion of N-acetylglucosamine-6-phosphate and N-acetylglucosamine-1phosphate (Cabib, et al., 1982). This enzyme is needed for the synthesis of the nucleotide sugar UDP-GlcNAc, which supplies the amino sugars for asparagine-linked oligosaccharides of glycoproteins. The N-glycosylation of proteins is essential for mitotic growth (Herscovics and Orlean, 1993). In addition, earlier demonstrations of a proliferative response of endothelial cells to Epo treatment (Anagnostou, et al., 1990; Carlini et al., 1995) strongly suggest that AGM1 is involved in angiogenesis. The signaling cascades that follow Epo interaction with receptors on erythroid cells are well documented (Ihle, 1995 ), involving tyrosine phosphorylation of the cellular kinase JAK2 and activation of the transcription factor STAT5. However, little is known regarding the Epo signaling cascade in endothelial cells. Endothelial cells express EpoR, interact with Epo, and this interaction results in the increased production of at least six distinct proteins (Haller, et al., 1996). The upregulation of AGM1 transcript in response to Epo in cultured endothelial cells indicates that AGM1 may be involved in the intracellular signaling cascade. Alternatively, the increased AGM1 synthesis may be one result of such a cascade. In either case, it would be of interest to determine whether those proteins shown to be upregulated by Epo treatment of endothelial cells, together with AGM1, form a functional ensemble that can be more completely understood and analyzed. The ability of endothelial cells to form capillary-like tubes is regulated not only by specific cytokine/receptor combinations, but also by the extracellular matrix. Several kinds of molecules on the endothelial cell surface act together to mediate cell–extracellular matrix interaction, including proteoglycans as well as proteins. The

best studied family of receptors that mediate cell–matrix interactions is the integrins, whose members serve an information-transfer function (Schwartz, et al., 1995). Integrin–ligand binding triggers cytoskeletal reorganization at specific sites on the surface membrane to facilitate cell movement, as well as activating intracellular pathways that can result in cell proliferation. One subfamily of integrins, the selectins, recognize as their binding site fucosylated glycans. Synthesis of these glycans includes an ordered series of posttranslational modifications, involving various nucleotide-sugar donors (Maly et al., 1996). Another possible function of increased levels of AGM1 could be to increase the size of the pool of UDP-GlcNAc for synthesis of glycosylated ligands related to integrin binding. The tissue distribution of AGM1 expression is compatible with the different levels of production of glycosylated proteins by these tissues. Thus, the expression is elevated in the pancreas, whose secretory functions are well known, and is very low in skeletal muscle (Fig. 4A). Since both skeletal muscle and heart tissues are composed largely of striated muscle, their difference in AGM1 mRNA levels very likely reflects the greater amount of endothelial cells in heart. The different levels of AGM1 expression in various adult tissues are in agreement with a similar tissue distribution of human phosphomannomutase expression (Matthijs, et al., 1997). In human endothelial cells and various adult tissues examined, the major transcript noted by Northern blot analysis was ~2.4 kb, sufficient to encode a 60 kDa protein. The reasons for the presence of larger transcripts in most of the adult tissues are not clear. However, mRNAs with extended 3∞- and 5∞- UNT are not uncommon, and these larger transcripts may have such UNT ends. However, the larger transcripts may encode other phosphoglucomutase isoenzymes. Future work will determine whether AGM1 is upregulated in vivo in brain capillaries or uterine tissue as a result of the paracrine action of Epo. If so, this locus may represent an important component in the mechanism involved in the control of blood vessel growth and development.

C. Li et al. / Gene 242 (2000) 97–103

Acknowledgements We wish to thank Andrea Molinaro of the Laboratory of Microchemistry for DNA sequencing and Tellervo Huima-Byron and Yelena Oksov for preparation of illustrations.

References Anagnostou, A., Lee, E.S., Kessimian, N., Levinson, R., Steiner, M., 1990. Erythropoietin has a mitogenic and positive chemotactic effect on endothelial cells. Proc. Natl. Acad. Sci. USA 87, 5978–5982. Anagnostou, A., Liu, Z., Steiner, M., Chin, K., Lee, E.S., Kessimian, N., Noguchi, C.T., 1994. Erythropoietin receptor mRNA expression in human endothelial cells. Proc. Natl. Acad. Sci. USA 91, 3974–3978. Bazan, J.F., 1990. Structural design and molecular evolution of a cytokine receptor superfamily. Proc. Natl. Acad. Sci. USA 87, 6934–6938. Bhattacharyya, N., Chattapadhyay, R., Hirsch, A., Banerjee, D., 1991. Isolation, characterization and sequencing of the chicken apolipoprotein-encoding gene. Gene 104, 163–168. Cabib, E., Roberts, R., Bowers, B., 1982. Synthesis of the yeast cell wall and its regulation. Annu. Rev. Biochem. 51, 763–793. Carlini, R.G., Reyes, A.A., Rothstein, M., 1995. Recombinant human erythropoietin stimulates angiogenesis in vitro. Kidney Int. 47, 740–745. D’Andrea, A.D., Lodish, H.F., Wong, G.G., 1989. Expression cloning of the murine eythropoietin receptor. Cell 57, 277–285. Digicaylioglu, M., Bichet, S., Marti, H.H., Wenger, R.H., Rivas, L.A., Bauer, C., Gassmann, M., 1995. Localization of specific erythropoietin binding sites in defined areas of the mouse brain. Proc. Natl. Acad. Sci. USA 92, 3717–3720. Haller, H., Christel, C., Dannenburg, L., Thiele, P., Lindschau, C., Luft, F.C., 1996. Signal transduction of erythropoietin in endothelial cells. Kidney Int. 50, 481–488. Herscovics, A., Orlean, P., 1993. Glycoprotein biosynthesis in yeast. FASEB J. 7, 540–550. Ihle, J.N., 1995. Cytokine receptor signalling. Nature 377, 591–594. Ingber, D.E., Folkman, J., 1989. Mechanochemical switching between growth and differentiation during fibroblast growth factor-stimulated angiogenesis in vitro: role of extracellular matrix. J. Cell Biol. 109, 317–330. Jelkmann, W., 1992. Erythropoietin: structure, control of production, and function. Physiol. Rev. 72, 449–489. Koury, M.J., Bondurant, M.C., 1992. The molecular mechanism of erythropoietin action. Eur. J. Biochem. 210, 649–663.

103

Kozak, M., 1991. Structural features in eukaryotic mRNAs that modulate the initiation of translation. J. Biol. Chem. 266, 19867–19870. Kyte, J., Doolittle, R.F., 1982. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 157, 105–132. Liang, P., Pardee, A.B., 1992. Differential display of eukaryotic messenger RNA by means of polymerase chain reaction. Science 257, 967–971. Maciag, T., Kadish, J., Wilkins, L., Stemerman, M.B., Weinstein, R., 1982. Organizational behavior of human umbilical vein endothelial cells. J. Cell Biol. 94, 511–520. Maly, P., Thall, A.D., Petryniak, B., Rogers, C.E., Smith, P.L., Marks, R.M., Kelly, R.J., Gerstan, K.M., Cheng, G., Saunders, T.L., Camper, S.A., Camphausen, R.T., Sullivan, F.X., Isogai, Y., Hindsgaul, O., von Andrian, U.H., Lowe, J.B., 1996. The a(1,3)fucosyltransferase Fuc-TVII controls leukocyte trafficking through an essential role in L-, E- and P-selectin ligand biosynthesis. Cell 86, 643–653. Matthijs, G., Schollen, E., Pirard, M., Budarf, M.L., Schaftingen, E.V., Cassiman, J.J., 1997. PMM (PMM1), the human homologue of SEC53 or yeast phosphomannomutase, is localized on chromosome 22q13. Genomics 40, 41–47. Morishita, E., Narita, H., Nishida, M., Kawashima, N., Yamagishi, K., Masuda, S., Nagao, M., Hatta, H., Sasaki, R., 1996. Antierythropoietin receptor monoclonal antibody: epitope mapping, quantification of the soluble receptor and detection of the solubilized transmembrane receptor and the receptor expressing cells. Blood 88, 465–471. Morishita, E., Masuda, M., Nagao, M., Yasuda, Y., Sasaki, R., 1997. Erythropoietin receptor is expressed in rat hippocampal and cerebral cortical neurons, and erythropoietin prevents in vitro glutamate-induced neuronal death. Neuroscience 76, 105–116. Orlic, D., Anderson, S., Biesecker, L.G., Sorrentino, B.P., Bodine, D.M., 1995. Pluripotent hematopoietic stem cells contain high levels of mRNA for c-kit, GATA-2, p45, NF-E2, and c-myb and low levels or no mRNA for c-fms and the receptors for granulocyte colony-stimulating factor and interleukins 5 and 7. Proc. Natl. Acad. Sci. USA 92, 4601–4605. Risau, W., 1995. Differentiation of endothelium. FASEB J. 9, 926–933. Risau, W., 1997. Mechanisms of angiogenesis. Nature 386, 671–674. Schwartz, M.A., Schaller, M.D., Ginsburg, M.H., 1995. Integrins: emerging paradigms of signal transduction. Annu. Rev. Cell Dev. Biol. 11, 549–599. Shaw, G., Kamen, R., 1986. A conserved AU sequence from the 3∞ untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 46, 659–667. Yamaji, R., Okada, T., Moriya, M., Naito, M., Tsuruo, T., Miyatake, K., Nakano, Y., 1996. Brain capillary endothelial cells express two forms of erythropoietin receptor mRNA. Eur. J. Biochem. 239, 494–500. Yasuda, Y., Masuda, S., Chikuma, M., Inoue, K., Nagao, M., Sasaki, R., 1998. Estrogen-dependent production of erythropoietin in uterus and its implication in uterine angiogenesis. J. Biol. Chem. 273, 25381–25387.